Efficacy of EGFR Tyrosine Kinase Inhibitors in Patients With EGFR-Mutated Non–Small-Cell Lung Cancer: A Meta-Analysis of 13 Randomized Trials

Review

Efficacy of EGFR Tyrosine Kinase Inhibitors in Patients With EGFR-Mutated Non–SmallCell Lung Cancer: A Meta-Analysis of 13 Randomized Trials Fausto Petrelli, Karen Borgonovo, Mary Cabiddu, Sandro Barni Abstract Advanced non–small-cell lung cancer (NSCLC) harboring activating mutations of epidermal growth factor receptor (EGFR) are particularly sensitive to tyrosine kinase inhibitors (TKIs), namely erlotinib and gefitinib. The purpose of this metaanalysis was to evaluate the benefit of EGFR TKIs in EGFR-mutated NSCLCs. Eligible studies included published randomized controlled trials in which erlotinib or gefitinib (alone or with chemotherapy) were compared with standard therapy in 1260 patients with EGFR-mutated NSCLCs who were included in 13 trials. The mutational status was obtained through a retrospective or prospective analysis. Relative risk (RR) was calculated for response rate, and hazard ratios (HRs) were calculated for progression-free and overall survival. EGFR TKIs increase the chance of obtaining an objective response almost 2-fold when compared with chemotherapy (RR, 2.06; 2p ⬍ .00001). The response rate was 70% vs. 33.2% in first-line trials. In 3 second-line trials, response rates were 47.4% vs. 28.5%, with a benefit similar to first-line trials (RR, 1.79; 2p ⫽ .04). EGFR TKIs reduced the hazard of progression by 70% in all trials (HR, 0.30; 2p ⬍ .00001) and by 65% in first-line trials only (HR, 0.35; 2p ⬍ .00001). Overall, however, they do not improve survival (HR, 0.96; 2p ⫽ .71). NSCLCs harboring EGFR mutations derive greater benefit from erlotinib or gefitinib than from chemotherapy. All patients affected by NSCLC with an EGFR-positive mutation test result must be offered the opportunity to be treated with an EGFR TKI upfront or during the natural course of the disease if not previously exposed. Clinical Lung Cancer, Vol. 13, No. 2, 107-14 © 2012 Elsevier Inc. All rights reserved. Keywords: EGFR, Erlotinib, Gefitinib, Mutation, Non–small-cell lung cancer

Introduction Although the small-molecule epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) erlotinib and gefitinib have modest clinical benefits after platinum-based chemotherapy in unselected patients with non–small-cell lung cancer (NSCLC), an emerging and potentially more elegant strategy is to move these agents to the frontline setting for selected patients. DNA sequencing of tumors from multiple series of patients with dramatic responses to gefitinib, compared with patients without responses, revealed the presence of characteristic genetic mutations in the EGFR gene.1-3 The previously identified clinical markers of response to EGFR TKIs Azienda Ospedaliera Treviglio-Caravaggio, Unità operativa di Oncologia, Treviglio (BG), Italy Submitted: Apr 22, 2011; Revised: Aug 26, 2011; Accepted: Aug 29, 2011 Corresponding author: Fausto Petrelli, MD, Azienda Ospedaliera TreviglioCaravaggio, Treviglio (BG), Italy Fax: ⫹39-0363424380; e-mail contact: [email protected]

1525-7304/$ - see frontmatter © 2012 Elsevier Inc. All rights reserved. doi: 10.1016/j.cllc.2011.08.005

were found to be commonly associated with the presence of these mutations; thus these clinical features are actually believed to be surrogates for the molecular biomarker of the EGFR mutation. The strong connection between response to EGFR TKIs and the presence of EGFR mutations was confirmed in a recent systematic review of 59 studies using first-line TKI treatment that included responses stratified by EGFR mutation status. This meta-analysis demonstrated that EGFR mutations predict response to EGFR TKIs, with a sensitivity of 0.78 (95% confidence interval [CI], 0.740.82).4 Many individual studies have also demonstrated better progression-free survival (PFS) and overall survival (OS) in patients with EGFR mutations that are treated with an EGFR TKI than in patients without such a mutation, suggesting that response to TKI treatment may also correspond to an improvement in survival.5-10 In particular, prospective clinical trials enriched with patients with EGFR-mutant NSCLC have consistently shown that gefitinib is active in this population.11-17 More than 90% of EGFR tyrosine kinase domain mutations associated with sensitivity to EGFR TKI therapy fall into 2 categories:

Clinical Lung Cancer March 2012

107

Erlotinib and Gefitinib in EGFR-Mutated NSCLC in-frame deletions in exon 19 and the L858R point mutation in exon 21, as reviewed by Sharma et al.18 Given the molecular biological characteristics of EGFR-mutant NSCLC, its particular sensitivity to TKIs, and the improved toxicity profile of TKIs compared with standard chemotherapy, a number of recent trials have been enriched with patients with EGFR-mutated tumors; these patients are more likely to benefit from first-line TKI treatment. Some have used EGFR mutations to identify patients, whereas others have used clinical criteria associated with response to TKIs—such as histologic type, smoking status, sex, and ethnicity; others have selected patients with potential intolerance to chemotherapy based on performance status and age. As described further on, EGFR mutation status has emerged as the most important factor predictive of the benefit of first-line EGFR TKI therapy. In first-line settings, trials using a purely clinical selection of patients who would be expected to be intolerant of chemotherapy or carriers of EGFR mutations have not demonstrated superiority of TKIs over standard care. The recent IPASS (IRESSA Pan-ASia Study) from Asia suggests that clinical selection of patients alone is inadequate and establishes a rationale for first-line TKI treatment of patients with EGFR mutations.19 In this study, 1217 Asian patients with NSCLC of adenocarcinoma histologic type who never smoked or were former light smokers were randomized to receive gefitinib or carboplatin and paclitaxel chemotherapy. The trial demonstrated that gefitinib treatment was superior to chemotherapy for PFS (HR, 0.74; 95% CI, 0.65-0.85; 2p ⬍ .001) in this clinically selected group of patients. In addition, patients who received gefitinib had fewer treatment-related side effects and an improved quality of life. Other recently published randomized studies consistently show a significant PFS advantage with initial TKI therapy instead of chemotherapy in patients with known EGFR mutations. As a consequence of the release of the results of this trial, European health authorities have approved gefitinib for the treatment of adult patients with locally advanced or metastatic NSCLC with activating mutations of EGFR.20 The aim of this meta-analysis was to compare efficacy of EGFR TKIs and chemotherapy (or placebo or best supportive care) in patient carriers of EGFR mutations enrolled in randomized trials.

Patients and Methods Search Strategy We performed a computerized search of MEDLINE, The Cochrane Central Register of Controlled Trials (CENTRAL), and the EMBASE database, using combinations of the following terms: gefitinib [Substance Name], or erlotinib [Substance Name], and Carcinoma, non–small-cell lung [MeSH] and mutation [MeSH] and Receptor, Epidermal Growth Factor or EGFR [MeSH]. The proceedings from the 2000-2011 conferences of the American Society of Clinical Oncology, European Society of Medical Oncology, and International Society for the Study of Lung Cancer World Conference of Lung Cancer were also searched for relevant abstracts (they had to contain a confirmatory update of already presented data) through EMBASE. No language limits were applied. The deadline for the trial inclusion was August 2011.

108

Clinical Lung Cancer March 2012

Trial Identification Criteria We considered all phase II/III randomized controlled trials in which previously untreated or pretreated patients with advanced/ metastatic NSCLC were prospectively randomized to receive gefitinib or erlotinib (experimental arms) vs. chemotherapy, placebo, or best supportive care (control arms). In particular, a meta-analysis was performed only for trials that reported data of patients for whom EGFR status was known and evaluated as mutated (either prospectively or retrospectively).

Outcome Definition The meta-analysis was conducted in order to detect significant differences in outcomes (as described further on) in the subpopulation of patients carrying an activating EGFR mutation (mainly exon 19 deletions or exon 21 point mutations) who were treated with erlotinib or gefitinib either in the first-line setting or in subsequent treatment settings. Primary outcomes for this meta-analysis were objective response rate, PFS, and OS.

Data Extraction and Management The following information was sought from each article, although some articles did not contain all the information: first author, year of publication, number of patients (total and EGFR mutation carriers), performance status, median age, chemotherapy regimens, rate of treatment crossover between arms, number of patients who obtained a response, HRs for OS and PFS and their 95% CIs.21

Statistical Analysis Trials had to report response rate or number of responses/number patients (for both control and experimental arms) and HRs for PFS and/or OS. The meta-analysis of relative risks (RRs) for the objective responses to treatment and the meta-analysis of HRs for OS and PFS were calculated using Review Manager (RevMan), version 5.1 (The Nordic Cochrane Centre, Copenhagen, Denmark) (The Cochrane Collaboration, 2008). Intention-to-treat analysis was performed for RR, OS, and PFS. A statistical test with a p value ⬍ .05 was considered significant. RR ⬎ 1 reflects a better overall response rate in the experimental arm, HR ⬍ 1 reflects fewer deaths or progression in the experimental arm and vice versa. In each meta-analysis, Cochran’s Q statistic and I2 statistics were calculated first in order to assess the heterogeneity among the proportions of the included trials. In case the p value was found to be ⬍ .1, the assumption of homogeneity was deemed invalid and the random-effects model was reported. Otherwise, the fixed-effects model was reported. All p values were 2-sided. All CIs had a 2-sided probability coverage of 95%. Analysis of subgroups was also performed with RevMan to explore any heterogeneous results (if any) and to investigate any differences between experimental drugs (ie, different activity of erlotinib vs. gefitinib in terms of response rate and PFS) and different treatment settings in patients (ie, different benefit of erlotinib and gefitinib as first- or second-line treatment) in relation to the outcomes. The procedure of subgroup analysis consisted of undertaking a standard test (test for subgroup differences) for heterogeneity across subgroup results. In particular an I2 statistic was also computed for subgroup differences. This value describes the percentages of variability be-

Fausto Petrelli et al Figure 1 Funnel Plot of Standard Error by Log-Hazard Ratio for Progression-Free Survival

Funnel Plot of Standard Error by Log hazard ratio 0.0

Standard Error

0.1 0.2 0.3 0.4 0.5 0.6 –3

–2

–1 0 1 Log hazard ratio

2

All were 2-arm studies. Overall, 8 trials included gefitinib arms (250 mg/day in 8 studies and 500 mg day in 1 study)22-24,27,31 and 5 erlotinib arms (150 mg/day).19,25,26,28,29,33,45 All trials enrolled patients with advanced NSCLC (almost all stage IV disease). Four trials prospectively randomized an Asiatic22,24,29 and a European45 population of only patients with EGFR-mutated NSCLC; conversely, the other 9 trials enrolled patients with a mixed population (EGFR mutations or not) of patients from Western countries, and the analysis of outcome according to mutation status was performed retrospectively and published separately.23,25-27,31,33 [19/ 28] In all studies, the EGFR mutation status was confirmed in 1260 patients (609 randomized to EGFR TKI arms and 567 to control arms [the randomization arm was unknown in the remaining studies]). Complete characteristics of included trials are shown in Table 1.

Response Rate

tween the subgroups that is due to a true statistical difference rather than simply a chance.

Publication Bias A number of steps were included in the study design to minimize the potential for publication bias. The search strategy was extensive. In addition, publication bias for PFS analysis (9 trials) was of borderline relevance according to the funnel plot, Begg and Mazumdar correlation test (2p ⫽ .08677), and the Egger test (2p ⫽ .04677) (Figure 1).

Results Selection of Trials The publications that were initially identified totaled 1229, of which 1216 trials were subsequently considered ineligible for different reasons and 14 were included in this meta-analysis.19,22-33,45 Among these 14 publications, 7 were in abstract form only; 1 of them28 also was an update of a previous full-text published article19 and another 131 collected and pooled the results of EGFR mutations in patients from 4 previously published randomized trials.19,23,46,47 One full-text publication27 collected data from 4 trials (2 were excluded, however, because they were prospective single-arm phase II trials; the 2 pooled INTACT 1-2 (Iressa NSCLC Trial Assessing Combination Treatment) studies instead were included in this analysis. Overall the search retrieved data from 13 randomized controlled trials. Finally, 2 abstract-only articles30,32 were excluded because 1 represented the first preliminary analysis (abstract only) of an unpublished trial32 and the other included another biological agent (bevacizumab).30

Characteristics of the 13 Randomized Trials in the Meta-Analysis All were phase III trials. Overall, 8 were first-line trials,19,22,24,27,28,29,33,45 1 was a maintenance trial,25 and 4 were second- (or third-) line trials.23,26,46,47 Overall, the population randomized in these 13 trials included 10,433 patients with NSCLC.

Data for objective response rate was available in 11 trials (3 firstline19,22,24,27,29,33,45 and 3 second-line studies23,31 (V15-32 and ISEL [IRESSA Survival Evaluation in Lung cancer] trials). Overall, RR for obtaining a response in EGFR-mutated NSCLC is 2-fold higher with EGFR TKIs than with chemotherapy (RR, 2.06; CI, 1.66-2.56; 2p ⬍ .00001; 2p for heterogeneity ⫽ .04; I2 ⫽ 48% according to random-effects model) (Figure 2A). The response rate was 67.6 vs. 32.8% in the 2 comparison arms (70% vs. 33.2% in the first-line setting only; RR, 2.14). In 3 second-line trials, response rates were 47.4 vs. 28.5%31 with a magnitude of benefit similar to that in first-line trials (RR, 1.79; 95% CI, 1.04-3.09; 2p ⫽ .04; fixed-effects model) (Figure 2B). The test for subgroup differences between first- and second-line trials was not significant, with no heterogeneity between these trials (␹2 ⫽ 0.40, df ⫽ 1 (2p ⫽ .53), I2 ⫽ 0%). This means that the increased benefit of anti-EGFR therapies compared with standard therapies (control arms) is formally and statistically maintained both in untreated and pretreated patients. Two trials did not report the rate of response in the EGFRmutated population.25,26 The meta-analysis of RR for erlotinib and gefitinib trials led to a pooled HR of 2.78 (2p ⬍ .0001) and 1.83 (2p ⬍ .00001), respectively. The test for subgroup differences between gefitinib and erlotinib trials was significant, with high heterogeneity between these trials (␹2 ⫽ 6.85; df ⫽ 1; 2p ⫽ .009; I2 ⫽ 85.4%). This means that erlotinib appears to have a greater effect on RR than does gefitinib.

Progression-Free Survival PFS data in patients with EGFR mutation was available in 9 trials (7 first-line,19,22,24,27,29,45 1 maintenance,25 and 1 secondline study.23 In 1 publication, HR for death but not for PFS was reported,26 and in the TRIBUTE trial, V-15-32 trial, and ISEL trial 31,33,46,47 HRs for PFS were not reported. Overall, the HR for PFS is 0.30 (95% CI, 0.22-0.42; 2p ⬍ .00001; heterogeneity, 2p ⫽ .0002; I2, 75% according to random-effects model) (Figure 3). The result is similar after including only first-line studies (RR, 0.35; 2p ⬍ .00001). The meta-analysis of PFS for erlotinib and gefitinib trials leads to a pooled HR of 0.19 (2p ⬍ .0001) and 0.39 (2p ⬍ .00001), respectively. The test for subgroup differences between gefitinib and erlotinib trials was significant, with high heterogeneity between these

Clinical Lung Cancer March 2012

109

Erlotinib and Gefitinib in EGFR-Mutated NSCLC Table 1 Characteristics of the 11 Randomized Trials Included in the Metanalysis Study author–year (ref.)

Trial N° enrolled ADK pts PS 0-1/ Histology median age (%)

Treatment arms

tot. EGFR mut. Crossover EGFR mut Response rate PFS mo OS mo (exp/ pts exp ⴙ % EGFR to TKI screened % exp/control (exp/control) control) control arms mut. 19-21 (%) pts RR (p) HR (p) HR (p) N° (%)

Mok TS–2009 (19) Yang CH–2010 (28)

IPASS 1217 90%/57

96,3%

A: Gefitinib 250 mg/day B: CBDCA AUC 5-6⫹Paclitaxel 200 mg/m2 BSA

39,5%

437

261 (59,7%)

96,1%

71,2%/47,3% RR 1.51 (p⬍0.001)

9,5/6,3 HR 0.48 (p⬍0.001)

mo N.A. HR 1.002 (p⫽0.990)

Maemondo M–2010 (22)

228 98,7%/63

93,4%

A: Gefitinib 250 mg/day B: CBDCA AUC 6 ⫹Paclitaxel 200 mg/m2 BSA

94,6%

228 (all enrolled pts)

228 (100%)

93,8%

73,7%/30,7% RR 2.4 (p⬍ 0.001)

10.8/5.4 HR 0.3 (p⬍0,001)

30.5/23.6 HR N.A. (p⫽0.31)

Douillard JY–2010 (23)

INTEREST 1466 88,4%/60,5

56,6%

A: Gefitinib 250 mg/day B: Docetaxel 75 mg/m2 BSA (2nd line)

37%

297

44 (15%)

86%

42,1%/21,1% RR 2 (p⫽0.04)

7/4.1 HR 0.16 (p⫽0.001)

14,2/16,6 HR 0,83 (p⫽0,59)

Mitsudomi T–2010 (24)

WJTOG3405 172 100%/64

83,5%

A: Gefitinib 250 mg/day B: Docetaxel 60 mg/m2 BSA– CDDP 80 mg/m2 BSA

59,3%

172 (all enrolled pts)

172 (100%)

100%

62,1%/32,2% RR 1.93 (n⫽117 with measurable disease) (p⬍0.0001)

9.2/6.3 HR 0.489 (p⬍0.0001)

N.A.

Cappuzzo F–2010 (25)

SATURN 889 100%/60

45,3%

A: Erlotinib 150 mg/day B: Placebo

67%

518

58 (11,1%)

84,4%

N.A.

mo N.A. HR 0.10 (p⬍ 0.0001)

mo N.A. HR 0.83 (p⫽0.6810)

Tsao MS–2005 (26)

BR.21 731 66%/61

50%

A: Erlotinib 150 mg/day B: Placebo

7,4%

177

40 (22,6%)

80%

N.A.

N.A.

mo N.A. HR 0.77 (p⫽0.54)

Bell DW–2005 (27)

INTACT 1 INTACT 2 2130 90%/60,6

52,3%

A: CDDP 80 mg/m2 BSA ⫹ GEM 1250 mg/m2 BSA ⫹/- Gefitinib 250 mg/day B: CBDCA AUC 6 ⫹ Paclitaxel 200 mg/m2 BSA ⫹/Gefitinib 500 mg/day

N.A.

312

32 (10%)

87,5%

72%/40% RR 1,81 (p⫽0,3)

6.7/4.5 HR 0.4 (p⫽N.A.)

mo N.A. HR 1.77 (p⫽N.A.)

Zhou C–2010 (29)

OPTIMAL 165 N.A./N.A.

87%

A: CBDCA AUC 5–GEM 1000 mg/m2 BSA B: Erlotinib 150 mg/day

N.A.

165 (all enrolled pts)

165 (100%)

91%

83%/36% RR 2.3 (p 0,0000)

13.1/4.6 HR 0.16 (p ⬍ 0.0001)

Kris MG–2009 (31)

ISEL 1692 66,5%/61,8

45%

A: Gefitinib 250 mg/day B: Placebo (pretreated)

3%

215

26 (12%)

82%

37.5%/0% RR N.A.

10.8/3.8 HR N.A.

N.A.

Maruyama R–2008 (46) Kris MG–2009 (31)

V 15-32 490 95,7%/56% ⬍64y

77,7%

A: Gefitinib 250 mg/day B: Docetaxel 60 mg/m2 BSA (2nd line)

53%

57

31 (54,4%)

96%

66.7%/45.4% RR N.A.

7.5/9.0 HR N.A.

N.A.

Eberhard DA–2005 (33)

TRIBUTE 1079 99,9%/62,6

61%

A: CBDCA AUC 6 ⫹Paclitaxel 200 mg/m2 BSA ⫹ Erlotinib 150 mg/day B: CBDCA AUC 6 ⫹Paclitaxel 200 mg/m2 BSA ⫹ Placebo

N.A.

228

29 (12,7%)

86,2%

53%/21% RR 2.5 (p⫽0,13)

N.A.

mo N.A. HR N.A. (p⫽0.96)

EURTAC 174/ 86%/ 66

N.A.

A: erlotinib 150 mg/day B: cisplatinum-based doublets

N.A.

1,227

174 (14.1%)

100%

58%/15% RR 3.89 (p⫽N.A.)

Rosell R (45)

5.2/9.7 HR 0.37 (p⬍0.0001)

N.A.

NA for updated analysis

Ref.: reference; n°⫽number; Pts⫽patients; PS⫽performance status; ADK⫽adenocarcinoma; TKIs⫽tyrosine kinase inhibitors; EGFR⫽epidermal growth factor receptor; mut⫽mutatated; RR⫽ risk ratio; PFS⫽progression free survival; OS⫽overall survival; mo⫽months; N.A.⫽data not available; CBDCA⫽carboplatin; CDDP⫽cisplatin; GEM⫽gemcitabine.

trials (␹2 ⫽ 2.19; df ⫽ 1; 2p ⫽ .09; I2 ⫽ 65.6%). This confirms the apparently greater progression-delaying effect of erlotinib when compared with gefitinib. Because only 1 second-line trial included reported PFS data, a subgroup analysis for line of treatment was not performed.

Overall Survival Data for median OS in patients with EGFR mutation were available from 6 trials (3 first-line,27,28 1 maintenance,25 and 2 secondline trials23,26). Overall, the HR for death is not significantly different when comparing EGFR TKIs and control arms (HR, 0.96; 95% CI, 0.76-1.21; 2p ⫽ .71; heterogeneity, 2p ⫽ 0.79; fixed-effects model). With a random-effects model, the results are exactly the same (HR, 0.96; 2p ⫽ .71) (Figure 4). Results for first-line– only trials are similar (HR, 0.99; 2p ⫽ .91).

110

Clinical Lung Cancer March 2012

Discussion The results of this meta-analysis, the first to our knowledge to collect data of all available EGFR-mutated NSCLCs enrolled in randomized trials (first- and second-line studies), confirms beyond all doubt that selecting patients with NSCLC for EGFR mutations and offering them an EGFR TKI results in a better response rate and progression-delaying effect than does standard chemotherapy. The performance appears similar in second-line settings in which the chance of obtaining a response is 63% higher with EGFR TKIs. A review of the literature presented by Kris et al in fact shows that the objective response rate is about 70% in both first- and mixed- (pretreated patients) line trials, even if it is slightly lower in non-Asian studies.31 Quite identical outcomes have recently been reported by Bria et al who performed meta-analysis on only first-line trials (except retro-

Fausto Petrelli et al Figure 2 (A) Forrest Plot for Relative Risk for Response Rate (all lines trials). (B) Forrest Plot for Relative Risk for Response Rate (first- and second-line trials)

A

B

Experimental Control Study or Subgroup Events Total Events Total Weight 7.3.2 meta-analysis of RR for response rate 3.3% 14 3 15 Eberhard 2005 8 Bell 2005 6.1% 9 4 23 17 Kris ISEL 2009 0.6% 5 0 21 8 Kris V 15-32 2009 8.0% 15 7 16 11 Mok IPASS 2009 129 20.6% 61 132 94 Douillard INTEREST 2009 4.6% 22 5 22 9 Maemondo 2010 114 17.3% 35 114 84 Zhou 2010 72 16.4% 26 82 68 Mitsudomi 2010 59 13.0% 19 58 36 Rosell EURTAC 2011 50 86 13 87 10.1% Subtotal (95% CI) 569 526 100.0% Total events 385 173 Heterogeneity: τ2 = 0.05; χ2 = 17.40, df = 9 (p = .04); I2 = 48% Test for overall effect: Z = 6.53 (p < .00001) Total (95% CI) 569 526 100.0% Total events 385 173 Heterogeneity: τ2 = 0.05; χ2 = 17.40, df = 9 (p = .04); I2 = 48% Test for overall effect: Z = 6.53 (p < .00001) Test for subgroup differences: Not applicable

Risk Ratio M-H, Random, 95% CI

Experimental Control Study or Subgroup Events Total Events Total 7.3.1 meta-analysis of RR for response rate (1st-line trials) Eberhard 2005 8 15 3 14 Bell 2005 17 23 4 9 Mok IPASS 2009 94 132 61 129 Maemondo 2010 84 114 35 114 Zhou 2010 68 82 26 72 Mitsudomi 2010 36 58 19 59 Rosell EURTAC 2011 50 86 13 87

Weight

Risk Ratio M-H, Fixed, 95% CI

Year

1.7% 3.2% 34.7% 19.7% 15.6% 10.6% 7.3%

2.49 [0.82, 7.55] 1.66 [0.77, 3.59] 1.51 [1.22, 1.86] 2.40 [1.78, 3.23] 2.30 [1.66, 3.17] 1.93 [1.26, 2.94] 3.89 [2.28, 6.63]

2005 2005 2009 2010 2010 2010 2011

484

92.7%

2.09 [ 1.82, 2.39]

22 15 5 42

2.8% 4.1% 0.4% 7.3%

1.80 [0.72, 4.52] 1.47 [0.78, 2.78] 4.64 [0.31, 69.37] 1.79 [ 1.04, 3.09]

526 100.0%

2.07 [ 1.81, 2.36]

Subtotal (95% CI) 510 Total events 357 Heterogeneity: χ2 = 16.11, df = 6 (p = .01); I2 = 63% Test for overall effect: Z = 10.52 (p < .00001)

2.49 [0.82, 7.55] 1.66 [0.77, 3.59] 4.64 [0.31, 69.37] 1.47 [0.78, 2.78] 1.51 [1.22, 1.86] 1.80 [0.72, 4.52] 2.40 [1.78, 3.23] 2.30 [1.66, 3.17] 1.93 [1.26, 2.94] 3.89 [2.28, 6.63] 2.06 [ 1.66, 2.56]

Year

Risk Ratio M-H, Random, 95% CI

2005 2005 2009 2009 2009 2009 2010 2010 2010 2011

2.06 [ 1.66, 2.56]

0.02

0.1 Favors control

1

10 50 Favors experimental

Risk Ratio M-H, Fixed, 95% CI

161

7.3.2 meta-analysis of RR for response rate (2nd-line trials) Douillard INTEREST 2009 9 22 5 Kris V 15-32 2009 11 16 7 Kris ISEL 2009 8 21 0 Subtotal (95% CI) 59 Total events 28 12 2 2 Heterogeneity: χ = 0.84, df = 2 (p = .66); I = 0% Test for overall effect: Z = 2.09 (p = .04) Total (95% CI) 569 Total events 385 173 2 2 Heterogeneity: χ = 17.40, df = 9 (p = .04); I = 48% Test for overall effect: Z = 10.69 (p < .00001) Test for subgroup differences: χ2 = .29, df = 1 (p = .59); I2 = 0%

2009 2009 2009

0.02

0.1 Favors control

1

10 50 Favors experimental

Abbreviation: CI ⫽ confidence interval.

spective analysis of 2 phase III trials27,33 included in our article).44 Our meta-analysis was also enriched with maintenance- and secondline studies in which outcome of EGFR-mutated NSCLCs was available. The amount of benefit, at least for response rate, seems equal to first-line trials. Furthermore, a survival gain remains to be demonstrated with up-front use of either erlotinib or gefitinib. EGFR mutation status is confirmed to be a strong predictor of response and PFS. However until now EGFR mutation did not appear to be a predictive factor for differential survival between EGFR

TKIs and chemotherapy. This may be due to the crossover effect in the population taking the EGFR TKI after progression on chemotherapy arms (⬎ 60% of patients in the Mok trial19 received gefitinib after chemotherapy failure). Postprogression treatments may have diluted the survival effect and make improvement of the OS endpoint a difficult outcome to obtain. However a PFS benefit that is so much larger may be due to the fact that about 25% of patients in 2 trials that enrolled only patients with EGFR-mutated NSCLC never received an EGFR TKI. In the clinical practice of a Western country

Clinical Lung Cancer March 2012

111

Erlotinib and Gefitinib in EGFR-Mutated NSCLC Figure 3 Forrest Plot for Hazard Ratio for Progression-Free Survival (PFS) (all trials)

Study or Subgroup

Log[Hazard Ratio]

SE

7.2.1 meta-analysis of HR for PFS 0.288 Bell 2005 –0.916 0.147 Mok 2009 –0.734 0.468 Cappuzzo 2010 –2.303 0.159 –1.204 Maemondo 2010 0.244 –1.833 Zhou 2010 0.582 –1.833 Douillard 2010 0.191 –0.715 Mitsudomi 2010 0.196 –0.994 Rosell EURTAC 2011 Subtotal (95% CI) Heterogeneity: τ2 = 0.15; χ2 = 28.21, df = 7 (P = 0.0002); I2 = 75% Test for overall effect: Z = 7.19 (P < 0.0001) Total (95% CI) Heterogeneity: τ2 = 0.15; χ2 = 28.21, df = 7 (P = 0.0002); I2 = 75% Test for overall effect: Z =7.19 (P < 0.0001) Test for subgroup differences: Not applicable

Weight

Hazard Ratio IV, Random, 95% CI

11.9% 16.2% 7.5% 15.9% 13.2% 5.6% 14.9% 14.7% 100.0%

0.40 [0.23, 0.70] 0.48 [0.36, 0.64] 0.10 [0.04, 0.25] 0.30 [0.22, 0.41] 0.16 [0.10, 0.26] 0.16 [0.05, 0.50] 0.49 [0.34, 0.71] 0.37 [0.25, 0.54] 0.30 [0.22, 0.42]

100.0%

Year

Hazard Ratio Stage IV, Random, 95% CI

2005 2009 2010 2010 2010 2010 2010 2011

0.30 [0.22, 0.42] 0.05 0.2 1 5 20 Favors experimental Favors control

Abbreviations: CI ⫽ confidence interval; HR ⫽ hazard ratio; SE ⫽ standard error.

Figure 4 Forrest Plot for Hazard Ratio for Overall Survival (OS)

Study or Subgroup

Log[Hazard Ratio]

7.1.2 meta-analysis of HR for OS Tsao 2005 Bell 2005 Cappuzzo 2010 Douillard 2010 Yang IPASS 2010

–0.261 0.571 –0.186 –0.186 0.002

Hazard Ratio IV, Fixed, 95% CI

SE

Weight

0.337 0.644 0.455 0.358 0.144

12.2% 3.3% 6.7% 10.8% 66.9%

0.77 [0.40, 1.49] 1.77 [0.50, 6.25] 0.83 [0.34, 2.03] 0.83 [0.41, 1.67] 1.00 [0.76, 1.33]

100.0%

0.96 [0.76, 1.21]

Subtotal (95% CI) Heterogeneity: χ2 = 1.68, df = 4 (P = 0.79); I2 = 0% Test for overall effect: Z = 0.37 (P = 0.71) Total (95% CI) Heterogeneity: χ2 = 1.68, df = 4 (P = 0.79); I2 = 0% Test for overall effect: Z = 0.37 (P = 0.71) Test for subgroup differences: Not applicable

100.0%

Year

Hazard Ratio Stage IV, Fixed, 95% CI

2005 2005 2010 2010 2010

0.96 [0.76, 1.21] 0.2 0.5 1 Favors experimental

2 5 Favors control

Abbreviations: CI ⫽ confidence interval; HR ⫽ hazard ratio; SE ⫽ standard error.

(and so not in a selected Asiatic population), the selection of patients with EGFR-mutated NSCLC works, as demonstrated by Rosell et al.34 In a similar European setting, however,35 the choice of erlotinib or chemotherapy or the reverse sequence in an unselected population (not screened for EGFR mutation) showed that median OS was superior in patients who received first-line chemotherapy compared with those who received first-line erlotinib (12.0 vs. 8.5 months; HR, 1.36; 95% CI, 1.12-1.65; 2p ⫽ .002). Overall, it can be presumed that the vast majority of the patients in this unselected population did not have an EGFR mutation and therefore never received the therapy that was likely to be most beneficial for them but conversely received erlotinib with a detrimental effect.

112

Clinical Lung Cancer March 2012

The main question is if treatment must start with an EGFR TKI or with chemotherapy in a patient with EGFR-mutated NSCLC. The response to this question is unknown at the moment. The only strategy that is neither useful nor approved is giving EGFR TKIs concomitantly with chemotherapy, even if the Bell analysis of the INTACT trials (platinum-based chemotherapy ⫾ gefitinib) documented a benefit (except for survival) for gefitinib plus chemotherapy in a limited number of patients with the EGFR mutation (n ⫽ 32). An increase in response rate is also observed with erlotinib when added to chemotherapy in a similar first-line setting.33 The choice of an EGFR TKI in a first-line setting has now, without any doubt, produced a strong level of evidence in patients who

Fausto Petrelli et al had positive test results for a mutation in their NSCLC specimens, both from a safety and efficacy (RR and PFS) point of view. The substantial minimal burden of toxicity of EGFR TKIs (mainly cutaneous and gastrointestinal effects) should especially lead to a preference for these agents in patients with poor performance status or contraindications for chemotherapy or in elderly patients who are unable to tolerate platinum agents. At the same time, a gross tumor burden should also be a criterion of choice for EGFR TKIs, considering the higher ORR achievable (about 70% across all first-line trials). Recently the EURTAC (European Erlotinib Versus Chemotherapy) study was released and the results for RR and PFS were confirmed at the 2011 International Society for the Study of Lung Cancer World Conference of Lung Cancer. It confirmed for the first time in a European population of patients with EGFR-mutated NSCLCs that erlotinib improves PFS, with acceptable toxicity compared with platinum-based chemotherapy.45 Both maintenance- and second-line studies with erlotinib and gefitinib also confirm the quality of these strategies in patients who are pretreated with chemotherapy. At least 4 maintenance trials investigated the role of EGFR TKIs as maintenance therapy in unselected (not progressing) patients after induction platinum-based chemotherapy.25,36-38 In particular, all 4 of these trials clearly showed significant positive improvements in PFS with EGFR TKI maintenance therapy, with a dramatic HR of 0.10 for PFS in patients with EGFR mutations in Cappuzzo et al’s25 study even if there was no OS benefit. In this meta-analysis, the most robust data derived from a secondline study were from the Douillard analysis of the INTEREST trial.23 In this trial, EGFR mutation–positive patients had longer PFS (HR, 0.16; 95% CI, 0.05-0.49; 2p ⫽ .001) and higher ORR (42.1% vs. 21.1%; 2p ⫽ .04). Therefore this meta-analysis makes it clear that it is advisable for all patients carrying the EGFR mutation whose disease progresses after first-line platinum-based treatment to be offered gefitinib or erlotinib if they have not initially been exposed to these agents. In patients with unknown mutation status, however, TKI use is potentially advisable anyway in patients in whom clinical determinants are predictive of such mutations (eg, never- smoker, female sex, adenocarcinoma histologic type, and Asiatic race). Finally, understanding different EGFR mutation types is important when evaluating test results to determine whether a patient is suitable for first-line erlotinib therapy. The most common EGFR mutations are deletions in exon 19 (consisting of a loss of 9-24 base pairs) and point mutations in exon 21 (L858R, L861Q). Patients with exon 19 deletions have a median PFS of 14.6 months on firstline TKI therapy, whereas patients with the L858R mutation have a median PFS of 9.7 months.39 Mutations in exon 20, however, have been associated with resistance to TKI therapy,40,41 although it is not clear whether these mutations necessarily predict resistance to TKI therapy when found in untreated patient specimens. This remains a largely debated question. In 2 (gefitinib) first-line trials, however, no difference was found for RR and PFS between L858R and exon 19 mutation groups.22,24 A fascinating but controversial setting for the use of these agents is the postoperative and neoadjuvant phase. In stating the inferior outcome of chemotherapy in patients with the EGFR mutation, it is logical to approach mutation carriers with EGFR TKIs. Recently, interesting results with adjuvant TKIs have been published by US

authors.42 In the same manner, the higher response rate with these biological agents makes it natural to consider treatment of this type in locally advanced (nonresectable) NSCLCs. In a randomized phase III trial, however, Goss et al demonstrated that adjuvant gefitinib treatment after complete resection of early-stage NSCLC conferred neither disease-free survival nor OS advantage in the overall population.43 EGFR mutation status in particular was associated with an HR for survival of 1.06 (2p ⫽ .83). Results of ongoing trials are awaited before a targeted agent can be recommended in early disease.

Conclusion In conclusion, NSCLCs harboring EGFR mutations derive greater benefit from erlotinib or gefitinib than from chemotherapy, either in first-line or subsequent lines of therapy. These agents double the chance of an objective response and reduce the risk of progression by about 70% but do not increase OS. These results are likely to be influenced by crossover treatments that formally abrogate any survival gain. The paradigm of up-front treatment in this setting has to be shifted from platinum-based chemotherapy to molecular targeted therapies. All patients affected by NSCLC with EGFR mutation– positive analysis in fact should be offered the opportunity to be treated with an EGFR TKI (according to the labeled indications) during the natural course of the disease.

Disclosure All authors state that they have no relevant relationships to disclose.

References 1. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004; 350:2129-39. 2. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 304:1497-500. 3. Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are common in lung cancers from “never smokers” and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 2004; 101:13306-11. 4. Dahabreh IJ, Linardou H, Siannis F, et al. Somatic EGFR mutation and gene copy gain as predictive biomarkers for response to tyrosine kinase inhibitors in non-small cell lung cancer. Clin Cancer Res 2010; 16:291-303. 5. Han SW, Kim TY, Hwang PG, et al. Predictive and prognostic impact of epidermal growth factor receptor mutation in non-small-cell lung cancer patients treated with gefitinib. J Clin Oncol 2005; 23:2493-501. 6. Cappuzzo F, Hirsch FR, Rossi E, et al. Epidermal growth factor receptor gene and protein and gefitinib sensitivity in non-small-cell lung cancer. J Natl Cancer Inst 2005; 97:643-55. 7. Chou TY, Chiu CH, Li LH, et al. Mutation in the tyrosine kinase domain of epidermal growth factor receptor is a predictive and prognostic factor for gefitinib treatment in patients with non-small cell lung cancer. Clin Cancer Res 2005; 11: 3750-7. 8. Cortes-Funes H, Gomez C, Rosell R, et al. Epidermal growth factor receptor activating mutations in Spanish gefitinib-treated non-small-cell lung cancer patients. Ann Oncol 2005; 16:1081-6. 9. Zhang XT, Li LY, Mu XL, et al. The EGFR mutation and its correlation with response of gefitinib in previously treated Chinese patients with advanced nonsmallcell lung cancer. Ann Oncol 2005; 16:1334-42. 10. Takano T, Ohe Y, Sakamoto H, et al. Epidermal growth factor receptor gene mutations and increased copy numbers predict gefitinib sensitivity in patients with recurrent non-small-cell lung cancer. J Clin Oncol 2005; 23:6829-37. 11. Inoue A, Suzuki T, Fukuhara T, et al. Prospective phase II study of gefitinib for chemotherapy-naïve patients with advanced non-small-cell lung cancer with epidermal growth factor receptor gene mutations. J Clin Oncol 2006; 24:3340-6. 12. Asahina H, Yamazaki K, Kinoshita I, et al. A phase II trial of gefitinib as first-line therapy for advanced non-small cell lung cancer with epidermal growth factor receptor mutations. Br J Cancer 2006; 95:998-1004. 13. Sutani A, Nagai Y, Udagawa K, et al. Gefitinib for non-small-cell lung cancer patients with epidermal growth factor receptor gene mutations screened by peptide nucleic acid-locked nucleic acid PCR clamp. Br J Cancer 2006; 95:1483-9. 14. Yoshida K, Yatabe Y, Park JY, et al. Prospective validation for prediction of gefitinib sensitivity by epidermal growth factor receptor gene mutation in patients with non-small cell lung cancer. J Thorac Oncol 2007; 2:22-8.

Clinical Lung Cancer March 2012

113

Erlotinib and Gefitinib in EGFR-Mutated NSCLC 15. Sunaga N, Tomizawa Y, Yanagitani N, et al. Phase II prospective study of the efficacy of gefitinib for the treatment of stage III/IV non-small cell lung cancer with EGFR mutations, irrespective of previous chemotherapy. Lung Cancer 2007; 56: 383-9. 16. Tamura K, Okamoto I, Kashii T, et al. Multicentre prospective phase II trial of gefitinib for advanced non small cell lung cancer with epidermal growth factor receptor mutations: results of the West Japan Thoracic Oncology Group trial (WJTOG0403). Br J Cancer 2008; 98:907-14. 17. Sugio K, Uramoto H, Onitsuka T, et al. Prospective phase II study of gefitinib in non-small cell lung cancer with epidermal growth factor receptor gene mutations. Lung Cancer 2009; 64:314-8. 18. Sharma SV, Bell DW, Settleman J, et al. Epidermal growth factor receptor mutations in lung cancer. Nat Rev Cancer 2007; 7:169-81. 19. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009; 361:947-57. 20. European Medicines Agency. Available at: http://www.ema.europa.eu/ema/ index.jsp?curl⫽/pages/home/Home_Page.jsp. Accessed: September 22, 2011. 21. Parmar MK, Torri V, Stewart L. Extracting summary statistics to perform metaanalyses of the published literature for survival endpoints. Stat Med 1998; 17: 2815-34. 22. Maemondo M, Inoue A, Kobayashi K, et al. Gefitinib or chemotherapy for nonsmall-cell lung cancer with mutated EGFR. N Engl J Med 2010; 362:2380-8. 23. Douillard JY, Shepherd FA, Hirsh V, et al. Molecular predictors of outcome with gefitinib and docetaxel in previously treated non-small-cell lung cancer: data from the randomized phase III INTEREST trial. J Clin Oncol 2010; 28:744-52. 24. Mitsudomi T, Morita S, Yatabe Y, et al. Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): an open label, randomised phase 3 trial. Lancet Oncol 2010; 11:121-8. 25. Cappuzzo F, Ciuleanu T, Stelmakh L, et al. Erlotinib as maintenance treatment in advanced non-small-cell lung cancer: a multicentre, randomised, placebo-controlled phase 3 study. Lancet Oncol 2010; 11:521-9. 26. Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer—molecular and clinical predictors of outcome. N Engl J Med [published correction appears in N Engl J Med 2006; 355:1746]. N Engl J Med 2005; 353:133-44. 27. Bell DW, Lynch TJ, Haserlat SM, et al. Epidermal growth factor receptor mutations and gene amplification in non-small-cell lung cancer: molecular analysis of the IDEAL/INTACT gefitinib trials. J Clin Oncol 2005; 23:8081-92. 28. Yang CH, Fukuoka M, Mok TS, et al. Final overall survival (OS) results from a phase III, randomized, open-label, first-line study of gefitinib (G) v carboplatin/ paclitaxel (C/P) in clinically selected patients with advanced non-small cell lung cancer (NSCLC) in Asia (IPASS). Program and abstracts of the 35th European Society for Medical Oncology Congress; October 8-12, 2010; Milan, Italy. Abstract LBA2. 29. Zhou C, Wu YL, Chen G, et al. Efficacy results from the randomised phase III OPTIMAL (CTONG 0802) study comparing first-line erlotinib versus carboplatin (CBDCA) plus gemcitabine (gem), in Chinese advanced non-small-cell lung cancer (NSCLC) patients (pt) with EGFR activating mutations. Program and abstracts of the 35th European Society for Medical Oncology Congress; October 8-12, 2010; Milan, Italy (abstract LBA13). 30. Johnson B, Miller V, Amler L, et al. Biomarker evaluation in the randomized, double-blind, placebo-controlled, phase IIIb ATLAS Trial, comparing bevacizumab (B) therapy with or without erlotinib (E), after completion of chemotherapy with B for the treatment of locally-advance. EJC Supplements 2009; 7:5. 31. Kris M, Mok T, Kim E, et al. Response and progression-free survival in 1006 patients with known EGFR mutation status in phase III randomized trials of gefitinib in individuals with non-small cell lung cancer. EJC Supplements 2009; 7:505 (abstract O-9003). 32. Lee JS, Park K, Kim S-W, et al. A randomized phase III study of gefitinib (IRESSATM) versus standard chemotherapy (gemcitabine plus cisplatin) as a first-line treatment for never-smokers with advanced or metastatic adenocarcinoma of the

114

Clinical Lung Cancer March 2012

33.

34. 35.

36.

37.

38.

39.

40. 41. 42. 43.

44.

45.

46. 47.

lung. IASLC World Lung Conference on Lung Cancer. San Francisco, California, July 31-August 4, 2009 (abstract PRS.4). Eberhard DA, Johnson BE, Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic indicators in patients with non-small-cell lung cancer treated with chemotherapy alone and in combination with erlotinib. J Clin Oncol 2005; 23:5900-9. Rosell R, Moran T, Queralt C, et al, for the Spanish Lung Cancer Group: Screening for epidermal growth factor receptor mutations in lung cancer. N Engl J Med 2009; 361:958-67. Gridelli C, Ciardiello F, Feld R, et al. International multicenter randomized phase III study of first-line erlotinib (E) followed by second-line cisplatin plus gemcitabine (CG) versus first-line CG followed by second-line E in advanced non-small cell lung cancer (aNSCLC): the TORCH trial. J Clin Oncol 2010; 28(suppl):15. Available at:www.asco.org. Accessed: September 22, 2011. Miller VA, O’Connor P, Soh C, et al. A randomized, double-blind, placebo-controlled, phase IIIb trial (ATLAS) comparing bevacizumab (B) therapy with or without erlotinib (E) after completion of chemotherapy with B for first-line treatment of locally advanced, recurrent, or metastatic non-small cell lung cancer (NSCLC). J Clin Oncol 2010; 27(suppl):18 (abstract LBA8002). Available at www.asco.org. Accessed: September 22, 2011. Perol M, Chouaid C, Milleron BJ, et al. Maintenance with either gemcitabine or erlotinib versus observation with predefined second-line treatment after cisplatingemcitabine induction chemotherapy in advanced NSCLC: IFCT-GFPC 0502 phase III study. J Clin Oncol 28(suppl): 15 (abstract 7507). Gaafar RM, Surmont V, Scagliotti G, et al. A double-blind, randomized, placebocontrolled phase III intergroup study of gefitinib (G) in patients (pts) with advanced NSCLC, non-progressing after first-line platinum-based chemotherapy (EORTC 08021-ILCP 01/03). J Clin Oncol 28(suppl): 15 (abstract 7518). Jackman DM, Miller VA, Cioffredi LA, et al. Impact of epidermal growth factor receptor and KRAS mutations on clinical outcomes in previously untreated nonsmall-cell lung cancer patients: results of an online tumor registry of clinical trials. Clin Cancer Res 2009; 15:5267-73. Kobayashi S, Boggon TJ, Dayaram T, et al. EGFR mutation and resistance of non-small-cell lung cancer to gefitinib. N Engl J Med 2005; 352:786-92. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain. PLoS Med 2005; 2:e73. Janjigian YY, Park BJ, Zakowski MF, et al. Impact on disease-free survival of adjuvant erlotinib or gefitinib in patients with resected lung adenocarcinomas that harbor EGFR mutations. J Thorac Oncol 2011; 6:569-75. Goss JD, Lorimer I, Tsao MS, et al. A phase III randomized, double-blind, placebocontrolled trial of the epidermal growth factor receptor inhibitor gefitinb in completely resected stage IB-IIIA non-small cell lung cancer (NSCLC): NCIC CTG BR.19. 2010; J Clin Oncol 28(suppl):18s (abstract LBA7005). Bria E, Milella M, Cuppone F, et al. Outcome of advanced NSCLC patients harboring sensitizing EGFR mutations randomized to EGFR tyrosine kinase inhibitors or chemotherapy as first-line treatment: a meta-analysis. Ann Oncol 2011 Feb 16. [Epub ahead of print]. Rosell R, Gervais R, Vergnenegre A, et al. Erlotinib versus chemotherapy (CT) in advanced non-small cell lung cancer (NSCLC) patients (p) with epidermal growth factor receptor (EGFR) mutations: Interim results of the European Erlotinib Versus Chemotherapy (EURTAC) phase III randomized trial. 2011; J Clin Oncol 29(suppl): abstract 7503). Maruyama R, Nishiwaki Y, Tamura T, et al. Phase III study, V-15-32, of gefitinib versus docetaxel in previously treated Japanese patients with non-small-cell lung cancer. J Clin Oncol 2008; 26:4244-52. Thatcher N, Chang A, Parikh P, et al. Gefitinib plus best supportive care in previously treated patients with refractory advanced non-small-cell lung cancer: results from a randomised, placebo-controlled, multicentre study (Iressa Survival Evaluation in Lung Cancer). Lancet 2005; 366:1527-37.